Accuracy of GDP growth forecasts for transition countries: Ten years of forecasting assessed

The paper analyses the accuracy of GDP growth forecasts prepared by the EBRD for 25 transition countries between 1994 and 2004. It provides the first comprehensive statistical analysis of the accuracy of output growth forecasts for transition countries in central and eastern Europe and the former Soviet Republics. We find that EBRD forecasts are mostly unbiased and efficient, i.e., forecast errors are not autocorrelated and do not depend on the value of the forecasted variable. In addition, we show that forecast accuracy improves with progress in transition as well as with the expansion in the information domain. We have also identified the Russian crisis as the only clear structural break in the available time series. Finally, we show that the EBRD forecast accuracy for late within-year GDP forecasts is better than the forecast accuracy of other institutions by 0.4 percentage points.